Design And Implementation Of Sight-singing Evaluation System Based On IOS

Sight singing is one of the basic abilities of professional music education.It is a process in which learners sing against musical scores and convert musical notes into sight singing notes.Sight singing exercises mainly rely on professional teachers for one-to-one guidance,which is inefficient.At present,there are some sight-singing practice software on the market that can simply evaluate the pitch,but cannot provide evaluation services that are closer to the practice and examination environment.Therefore,the main work of this article is about this issue.The main contents of this article include: 1)A complete set of sight-singing evaluation solutions is proposed,which consists of three core algorithms: onset detection,note alignment,and pitch extraction.In terms of note start detection,onset detection algorithm based on CNN(Convolutional Neural Networks)is used;in the note alignment,a modeling method based on relative pitch is adopted,which is in the case of continuous wrong sounds in sight singing The accuracy is much higher than the note alignment method based on absolute pitch;in terms of pitch extraction,the existing MFSHS(Max-Frequency-SHS)algorithm of the research group is used;the entire program is built on the self-built 60-single test data set A note-level test was conducted,and its accuracy reached 90.3%.2)Design and implement a complete sight-singing evaluation system,which completely encapsulates the above solution into a highly concurrent back-end HTTP service,providing transparent data access and computing support for a diverse front-end user interaction subsystem.3)Designed and implemented a set of front-end interactive subsystems based on i OS Swift language.On the basis of providing a friendly interactive interface for sight-singing practitioners,more emphasis was placed on the system performance optimization design based on fragment uploading,and corresponding tests were conducted.discuss.This paper designs and implements a set of i OS-based audiovisual evaluation system,which won the second prize of the 2019 China University Computer Competition Mobile Application Innovation Competition.The system can provide users with real-time and accurate sight-singing practice services.